Sometimes, the PDF title isn't correct. We need to check that it is getting passed the correct string. It is possible this might be fixed when we fix the issue of the old datasets not being downloaded.